A semiotic theory of memory: between movement and form*

1 Introduction

Memory lives in its more or less provisional, semiotic materializations: texts, monuments, images, music, performances, rituals and daily interactions and practices. Its condition of existence lies both in its capacity to take a recognizable cultural form within a given context and to break away from that very context, in order to be transmitted across time, space and generations. Hence, when analyzing memory, we are asked to acknowledge its puzzling nature: memory is always localized and always displacing, living in its transmission from one context to another, between local mise en forme and movement.

However, within the multidisciplinary field of memory studies, remembering and forgetting have been examined through two methodological and epistemological models, which have been presented as counterposed or mutually exclusive: containment and flow. The former conceptualizes memory as contained within cultural artefacts, that is, within their symbolical, textual, discursive and even physical boundaries; the latter looks at memory as a migration of contents and forms from one cultural manifestation to another, taking place in a chain generated – over time – around a subject matter.

Contributing to a growing scholarship and discussion in the field of memory studies, this article argues for an integrated approach, capable of describing remembering and forgetting both as movement and form. It will do so by drawing on Umberto Eco’s encyclopedic and interpretative model. In the 2000s, Eco looked back at his own interpretative and encyclopedic theory, envisioning it also as a model for analysing and studying cultures in terms of memory-making. In so doing, he developed a perspective that is consistent with a line of investigation Jurij Lotman inaugurated back in the 1970s. In the essay “From the Tree to the Labyrinth,” which opens – and gives the title to – his last and most important collection of theoretical essays, Eco argues that memory, abeyance and oblivion are the result of processes of interpretation, which select and organize knowledge locally (2014 [2007]). In order to describe these processes, we have to take into consideration both semiosis – i.e., the production of the flow of interpretants – and those local conditions that select relevant knowledge, using the analytical toolbox elaborated in the semiotics of text. This transforms the interpretative semiotics model into a theoretical and methodological perspective, which allows us to combine two images of memory: always moving yet also localized.

Semiotics does not aim to differentiate between what is true and false, or to reconstruct facts and events per se; it does not study the “world” or “things” as they are, but the “n possible versions” we can give of these (Paolucci 2016: 112 my translation), and the ways in which they are connected and in relationship. Likewise, a semiotic perspective applied to memory does not aim to study the past per se, but the n possible versions and interpretations of the past that we can locally identify, and their mutual relationships. These versions can be either consistent or conflicting and contradictory, co-existent or successive, drawing on different semiotic substances (images, verbal and oral, rituals, etc.), media (cinema, TV, newspapers, etc.), and genres (novels, testimonies, historiography, etc.), each with its own formal rules for production and reception of cultural artefacts and of “forming the world.” Hence, we need a model capable of combining the form and the movement as well as being able to analyze the relationship between versions of the past that are produced by and through different media, substances, and genres.

I will develop my argument in three steps. First, I will describe the genealogy and the semiotic characteristics underpinning the memory-as-flow and memory-as-containment models. Second, I will identify the place of semiotics within this contraposition, integrating it into the broader field of memory studies and looking, in particular, at Eco’s semiotic theory. Third, I will describe an exemplary case study, capable of showing how memory-as-flow and memory-as-containment can be envisioned together, in a combined analysis of remembering and forgetting. The case study deals with the memory of the former concentration camp of Fossoli, an Italian location that served different functions between the 1940s and the 1970s. By tracing the use of the word “Fossoli” in Italian newspapers over a time period of 40 years, we will see how this polysemic and multilayered space of meaning makes sense over time, in the selection of knowledge that generates remembering and forgetting as effects of interpretation as well as multiple versions of what Fossoli is and means.

2 Memory studies: the disciplinary division of labor

In an overview of the memory studies field, Astrid Erll (2011) aptly described two methodological perspectives for conceptualising, investigating and understanding how we remember and forget: “framed-ness” and “travelling memory.” Framed-ness is about studying memory within social formations, for example “a religious group, a social class, and ethnicity” (Erll 2011: 6), and the result is the image of a “containered” memory (Erll 2011: 11). Erll traces this model back to Maurice Halbwachs’s idea of “memory frameworks,” whose most successful – though also contested (Gensburger 2016) – interpretation is to be located in Pierre Nora’s concept of lieu de mémoire. By contrast, the expression “travelling memory” is meant to stress the fact that “memory lives in and through its movements” and does not stand still (Erll 2011: 11); it therefore needs a model capable of describing such movements. This was the position of Aby Warburg (expressed in the idea of Nachleben) and – more recently – of Jan Assmann (through the concept of mnemohistory).

In his response to Erll’s survey of the memory studies field, Jeffrey Olick (2014) adds that the most recent theorisations in memory studies try to “transcend the ‘container’” (Olick 2014: 23), in particular that of the nation-state. Indeed, if we look at the metaphorical fields used in the most recent theories of memory studies, we can clearly see a shift in focus from the act of bordering – suggested by the words “framework” (Halbwachs) and “lieu” (Nora) – to the idea of movement and change: for example, in the concepts of “multidirectional” (Rothberg 2009) and “connective” memory (Hoskins 2011). According to Olick, such a shift matches, or is related to, developments in the new media environment, “which is characterized by fluidity, boundary-crossing and hybridity” (Olick 2014: 23).

More recently, Andrew Hoskins (2018) has offered an in-depth analysis of this shift by drawing on media history and, in particular, on a tripartition elaborated by Thomas Pettitt. The latter argued that, prior to the fifteenth century, in a world dominated by orality, culture and media initially worked through connections. Subsequently, after the rise of the printing press and up to the era of digital media, media environments were dominated by the model of containment, because memory was conceptualized as being contained within the boundaries of printed texts or (television) screens. Finally, in the world we currently inhabit, which is dominated by digital media, we have returned to the connection model. Hoskins argues that memory and oblivion nowadays happen in “the multitude of techniques, technologies and practices through which discourse and interaction is mediated. This is the entire ‘semiotic environment’ in which memory is understood and made relevant to a person, given community or group” (Hoskins 2018: 8). Hence, we need “a new ontology for memory studies,” capable of tracking memories from “representation to enfolding, from space to time, from distribution to hyperconnectivity” (Hoskins 2018: 7).

In this article, I will partially depart from the presentation of flow and containment as two historical and/or ontological, mutually exclusive models in memory- and oblivion-making. Indeed, the contraposition between flow and containment is to be understood at a methodological and epistemological level. Remembering and forgetting have, in fact, always emerged from a semiotic environment that connects a multitude of “techniques, technologies and practices,” which is not something that specifically marks our present. This is evident even in Halbwachs’s description of memory-making.

As Erll and Rigney have already stressed, although Halbwachs paid “only incidental attention to the role of media in memory-making” (Erll and Rigney 2009: 1), the French sociologist described memory as the result of an assemblage of different sources, genres and media. In La mémoire collective, Halbwachs describes a visit to London, where he ends up strolling through the British capital in the company of four different figures: an architect, who “directs my attention to the character and arrangement of city building”; a historian, who “tells me why a certain street is historically noteworthy”; a painter, who “alerts me to the colors in the parks”; and a businessman, who “takes me into the public thoroughfares” (Halbwachs 1980 [1950]: 23). During the visit, Halbwachs also makes use of a map, and the city reminds him “of Dickens’s novels read in childhood” (Halbwachs 1980 [1950]: 23). The discursive characterisation of his companions, and the references to maps and literature, is symptomatic of how – even for Halbwachs – memory is the result of the connection and assemblage of knowledge shaped by different discursive genres, media and social practices (here represented by his companions); the latter “do things with words” – and with the world – according to their specificities and social functions. Halbwachs’s companions and texts offer multiple and complementary versions of the city according to different cultural and discursive practices of forming the world: they look at London under different respects and with varying capacities, to use Charles Sanders Peirce’s words. However, rather than to focus on the media and discursive practices in which his companions have specialised (painting, historiography, architecture, etc.), Halbwachs prioritises the study of social interactions (e.g., conversations during a visit) and social groups (e.g., his companions’ different professional milieus) as frameworks of memory. This is due to Halbwachs’s sociological perspective.

Indeed, what is at stake in the contraposition between flow and containment is not the history (or the ontology) of memory and media, as Hoskins argues, but rather, the division of labour between disciplines, and the ways in which they enact their objects of analysis through specific methods of analysis. Memory studies is a multidisciplinary field of research, whose practitioners have originally been trained within their own disciplines, and who have often specialized in specific media, and in medium-specific methods of analysis. For example, literary scholars have predominantly looked at individual texts and at a singular medium, which they consider as frames and carriers of memory. Even scholars in media studies have mainly focused on a single medium rather than studied the articulation between media. This fact may methodologically hamper the ecological validity of analysis in memory studies as well as the possibility of putting different versions of the past in relationship to one another, regardless of the medium, genre and semiotic substance that is used.

More recently, Robin Wagner-Pacifici – from a critical standpoint towards the memory studies project – has identified a contraposition between “stability and movement” (2017: 12), in the construction of events and their transmission over time. Wagner-Pacifici points at the analytical limit in sociology to capture and describe the phenomena of stability (containment) and movement (flow) from both perspectives.

This article aims at proposing a model of memory- and oblivion-making capable of taking into account, and considering together, both aspects; these are not to be considered as mutually exclusive but as co-constitutive. In doing so, I join a body of literature that has been developing in memory studies over the last two decades. In the early 2000s, Ann Rigney argued for an analysis of literary texts both as products and as agents, coining the definition of “portable monuments, which can be carried over into new situations” (2004: 383).^[1] She suggests this is done by morphing and transforming their meanings and characteristics:

when the various approaches to literary works (as product, as agent) are taken together, then a double picture emerges of their role in cultural remembrance. Firstly, literary works resemble monuments in that they provide fixed points of reference . . . At the same time as they may enjoy this monumentality, however, literary works continuously morph into the many other cultural products that recall, adapt, and revise them in both overt and indirect ways. (Rigney 2008: 349)

Umberto Eco’s interpretative semiotics offers a methodological perspective on memory and oblivion, which makes it possible to bridge the “double picture” that emerges in the contraposition between memory-as-containment and memory-as-flow. Before dealing with this approach, I will describe more in depth the origins and characteristics of the containment and flow models, as they have emerged in the memory studies debate. This will allow me to locate interpretative and cultural semiotics within the field of the investigation of memory, and distinguish between these two perspectives on memory-making.

3 Memory-as-containment

The containment model is grounded in the work of Maurice Halbwachs. Halbwachs theorized the primacy of the collective and social over the individual: the subject always remembers within frameworks, in particular within social groups. According to Halbwachs, remembrance is always the result of the subject’s position within a structured ensemble of relations.

Halbwachs’s work is marked by a relational approach. Remembrance makes sense only within a system of relations and differences, in which the whole defines the identity and meaning of each individual element:

When we look in the sky for two stars belonging to different constellations, we readily imagine that by merely tracing an imaginary line between them we confer on them some sort of unity. Nevertheless, each is only an element in a group and we were able to recognize them because neither constellation was then hidden behind a cloud. Similarly, since two thoughts contrast and apparently reinforce one another when brought together, we think they form a self-existing whole, independent of their parent wholes. We fail to perceive that in reality we are considering the two groups simultaneously, but each from the viewpoint of the other. (Halbwachs 1980 [1950]: 41–42)

A pupil of Emile Durkheim, Halbwachs seems to envision a structuralist approach to the study of memory, thus taking part in the imminent rise of structuralism coined in terms of an interdisciplinary koine. From this perspective, what is important is not the element per se, but its position in a system of relations. Another element that Halbwachs shares with structuralism is his synchronic approach. Counterposing historical chronology to collective tradition, Halbwachs points to an important difference between history and sociology: chronological time is “not the time in which [the groups’] collective thought habitually functioned or localized what was remembered of their past” (Halbwachs 1980 [1950]: 106). Social groups live and remember in synchronic states.

Hence, frameworks are methodologically conceptualized in terms of structures, and the attention given to social groups, defined as frameworks for individual remembrance, is also the result of another element: Halbwachs investigated memory within the study of social classes and groups.

Pierre Nora partially follows the Halbwachsian perspective and, in particular, his preference for the synchronic over the diachronic, for structure, and for the study of memory as a tool for analyzing other cultural phenomena, especially collective identities.

The point of departure . . . was to study national feeling not in the traditional thematic or chronological manner but instead by analyzing the places in which the collective heritage of France was crystallized . . . the specific role that memory played in the construction of the French idea of the nation. (Nora 1998: xv–xvi)

The central point, the goal is to reinterpret the history of France in symbolic terms . . . a history that is interested in memory not as remembrance but as the overall structure of the past within the present. (Nora 1998: xxiv)

In Nora’s words, the objects of investigation are the nation-state (as opposed to social groups, for Halbwachs), national sentiment, and the role memory plays in it. Furthermore, Nora’s monumental collection is organised like a dictionary, with an entry for each “lieu,” which is conceptualized as a symbolical “container” of memory for the nation as a collective subject.

In brief, while Halbwachs prioritised social interactions – milieux of memory – as opposed to the materiality of objects, languages, genres and media, Nora focused his attention on cultural artefacts, considered as carriers and containers of memory – lieux of memory. He thus downplayed the role of social actors as bricoleurs that assemble and connect different media, genres and practices within a localized, social practice. In spite of the radical contraposition between milieu and lieu of memory, which questions Nora’s status as actual successor of the Halbwachsian approach (Gensburger 2016), both are interested in describing “structures,” and both took part – in different ways and at different times – in the dominant, epistemological koine that was twentieth-century structuralism.

Within the containment model and perspective, we can identify two key concepts of memory studies: memoryscape and collective memory. The concept of memory in terms of a scape, first of all, is already present in Halbwachs’s metaphor of the skyscape: memory is the result of a synchronic configuration, the co-presence of different elements whose meaning emerges from their mutual relationship and position within a network at a given time. Secondly, the concept of collective memory stems from the primary research interests of both Halbwachs and Nora, namely to study the making of identities – at the level of social groups for Halbwachs, at the macro-level of the nation for Nora.

To summarize, the containment model deals with memory in three ways: as a specific cultural practice that we study in order to investigate other phenomena, namely collective identities, in the case of Halbwachs and Nora, but also literature, cinema, radio, TV and so on, when this approach is adopted in other disciplines, which take memory as an epiphenomenon of their disciplinary object of analysis; as a network of relationships that form structures; as contained in social (i.e., classes, groups) or symbolical (i.e., texts, rituals) structures, conceptualized as a network of relationships.

4 Memory-as-flow

Erll has observed that the flow model must be identified with Aby Warburg’s work, in particular. As an art historian, Warburg urged scholars to stop policing disciplinary boundaries, calling for methods of investigation that could cross these boundaries. Although he did not offer an explicit theoretical and methodological reflection on how to study memory, his approach can be seen at work in particular in the unfinished work Mnemosyne:^[2] an atlas of images for the tracking of survivals and returns, or what he dubbed the Nachleben (differently translated as “survival,” “revival” or “afterlife”) of images, themes and pathos, together with the transformation of meanings. As Georges Didi-Huberman argues, Mnemosyne is memory at work, “memory as such, the ‘living’ memory, from which [Warburg] derived the proper name that was to be given to the whole enterprise: Mnemosyne, the classic personification of memory, mother of the nine Muses” (Didi-Huberman 2017 [2002]: 296). By remediating the past and the archive through photography, Warburg put together different genres, media and semiotic substances: newspaper reports, postcards, advertisements, architectural sketches, maps, paintings, sculptures, drawings, pages of books, and so on. Every panel is a multimedia and multi-genre space of meaning, but also a multi-temporal one; in the historical difference between one element and another, we can reconstruct a diachronic order that is offered, though, in a synoptic, syncretic and synchronic space. Using montage as a method, Warburg made patterns emerge in the comparison between images, in their differences and similarities, in the hand-offs of forms that – as virtualities – migrate from one image to another, underlying their manifestation and underpinning their intelligibility (Zucconi 2018: 129). Warburg revealed – through the anachronisms caused by the placement of images belonging to different historical periods (diachronic dimension) in the same space, synchronically and synoptically – the processes of migration and translation of forms, and their (morphological) persistence through their (semantic) transformations.

In sum, Warburg seems to suggest that memory is not just a specific practice that – along with other discursive and social practices – is part of a collectivity’s cultural life; memory and culture are actually two faces of the same coin, and cultures can be described as processes of survivals and returns.

In the 1990s, Jan Assmann coined the term mnemohistory, in reference to the study of “the vertical lines of transmission and reception,” which is directly linked to Warburg’s reconstruction of the pathways of cultural memory. Mnemohistory is not interested in the reconstruction of facts as they supposedly emerged, but in the way “facts” are interpreted, received and remembered:

mnemohistory is concerned not with the past as such, but only with the past as it is remembered. It surveys the story-lines of tradition, the webs of intertextuality, the diachronic continuities and discontinuities of reading the past . . . it has an approach of its own in that it deliberately leaves aside the synchronic aspects of what it is investigating. It concentrates exclusively on those aspects of significance and relevance which are the product of memory – that is, of a recourse to a past – and which appear only in the light of later readings. Mnemohistory is reception theory applied to history [where] the “truth” of a given memory lies not so much in its “factuality” as in its “actuality.” (Assmann 1997: 9)

This definition of mnemohistory is extremely important as it is already methodologically operative. Mnemohistory focuses on the diachronic chain of interpretations that an event generates, in the reception and interpretation of “facts” and “events” (i.e., actuality) more than in the very reconstruction of facts (i.e., factuality). Hence, a mnemohistorical investigation should reconstruct the webs of intertextuality as discourse, that is, as the “concatenation of texts which are based on each other and treat or negotiate a common subject matter” (Assmann 1997: 15).

According to Assmann, such a concatenation is “a kind of textual conversation or debate which might extend over generations and centuries, even millennia” (Assmann 1997: 15). In discourse, we have to analyze narrative structures that are “operative in the organisation of action, experience, memory, and representation” (Assmann 1997: 15), and examine how such narratives emerge, circulate and are interpreted forming (and being formed by) grand narratives that are underpinned by semantic oppositions (Assmann 1997: 7).

In sum, the flow model calls for the construction of various corpora of texts, which Assmann calls discourse, and which Warburg constructed as multimedia, multi-genre and multi-temporal space. It then traces the linkages and transformations of a given “subject matter” (for Assmann, who seems more focused on the plan of content), or of a given cultural form (for Warburg, who seems more focused on the plan of expression).

If we compare the flow and containment models (see Table 1), we could say that the flow model focuses on a diachronic perspective (more present in Assmann than in Warburg), as opposed to the containment model, which focuses on synchronic states. While the containment model supports the idea of memoryscapes (i.e., the study of a synchronic state of memory), the flow model supports the idea of afterlives: the study of transformations over time. Furthermore, while the containment model conceptualizes memory as a specific cultural practice among others, through which to study social phenomena (in particular collective identities and social groups, or specific media, genres, cultural artefacts, etc.), the flow model conceptualizes memory and culture as two co-existing phenomena. In the latter case, to study memory is actually to study culture and the ways in which it “survives” over time, subject to continuities and discontinuities, persistence and transformation. If, in the containment model, we talk about “collective memory,” because we study memory mainly for an understanding of identity-making processes, in the flow model we talk about “cultural memory,” because we analyze memory in order to understand cultural dynamics at large. To think of culture as memory also blurs the difference between intentional and non-intentional acts of remembrance: in this perspective, remembering and forgetting are two co-constitutive phenomena that are present in every act of our social and cultural way of being and doing, although at different degrees of awareness, formality and symbolical relevance.

The containment and flow models, with the characteristics identified above, are to be considered two ideal-typical models. As we have seen so far, nuances, differences and even intersections can be perceived even between the authors of reference of the two models.

Semiotics offers the possibility to envision containment and flow as complementary models: memory migrates, travels, moves but also stops, sediments, rests and crystallizes, assuming more defined and clear-cut forms. We therefore need a model capable of considering and capturing both moments. In semiotics, such a dichotomy is translated into the contrast between structure, defined in terms of a synchronic system of relations between elements, and semiosis, the ways in which meanings form and take shape through temporalised processes of interpretation.

In particular, semiotics as developed by Umberto Eco, consistent with a line of investigation that already emerged in Jurij Lotman’s works in the 1970s, can be considered as an attempt to reconcile these two ways of grasping cultural phenomena: the restless movement of semiosis, and the transitory stabilisation of meanings in more structured, cultural forms. This epistemological and methodological model can help us envision the puzzling nature of memory between movement and form, between flow and containment.

5 Memory and semiotic studies

Jan and Aleida Assmann (Assmann 2011 [1992]: 7; Tamm 2015a: 128) have acknowledged that the concept of cultural memory was inspired by the work of the semiotician Jurij Lotman, in collaboration with Boris Uspensky (Lotman and Uspensky 1985). Lotman defines “culture as the nonhereditary memory of the community” (Lotman and Uspenskij 1978 [1971]: 213). Culture and memory are thus envisioned as two co-existing phenomena; this characteristic is a feature of the flow model, as we have seen above.

Lotman imagined cultures as semiospheres: systems that, like the biosphere, articulate multiple levels and regulate different phenomena, including the relationship with the outside. The aim of the semiotics of culture is to describe the “functional correlations” (Uspensky et al. 1998 [1973]) between the different sign systems – like media, semiotic substances, cultural artefacts and discursive domains – that constitute semiospheres, as well as the rules that regulate their relationships. Indeed, according to Lotman, culture is the result of the restless processes of creation, recording, location, circulation, translation, (re)organization, abeyance, recovery and erasure of knowledge and information. Lotman describes these processes using two concepts, informative memory and creative memory (Lotman 2000 [1985]), which are very close to the ideas of storage and functional memory as elaborated by Jan and Aleida Assmann (2011 [1999]: 119–135):

Lotman’s distinction between informative and creative cultural memory is quite similar to the distinction made by Aleida and Jan Assmann between storage memory and functional memory, where the task of the former is to record a maximum amount of information in a culture, and of the latter, to create new connections, constellations of these messages, thereby providing them with a new meaning and actuality. (Tamm 2019: 11)

Since the 1970s, Umberto Eco has fostered the translation and circulation of Lotman’s work, in the Italian and subsequently also in the international academic environment. Eco argued that the Estonian semiotician applied the method of structuralism but “offering a more complex and articulated approach” (Eco 1990: x). Indeed, Eco and Lotman’s theoretical paths and attitudes resonated with each other; they shared a constructive critique of structuralism, opening structuralist methods to a more dynamic, articulated and less dogmatic approach. On the one hand, Lotman worked on the relationship between culture and memory, analyzing structures of texts in order to grasp the dynamics of cultures as well as the functional correlations between different semiotic systems; on the other hand, Eco developed a general theory of culture so as to describe the general movement and production of interpretations, drawing on Peirce’s philosophy of semiosis, although he also looked at how texts, images and, in general, any cultural artefact shape semiosis locally.

Eco elaborated a concept that has many points of contact with Lotman’s semiosphere, as well as with the concepts of informative and creative memory: the Encyclopedia^[3]. As we will see, the concepts of informative memory (Lotman) and storage memory (Aleida Assmann), considered as spaces where “to record a maximum amount of information,” are close to the idea of Eco’s Maximal Encyclopedia. Likewise, the idea of creative memory and functional memory is very close to Eco’s idea of local Encyclopedias, which are the results of a selection and actualization within local processes of interpretation of sets of information and knowledge that are stored in the Maximal Encyclopedia.

Mnemohistory’s theoretical foundation rests on hermeneutical philosophy and reception theory (Assmann 1997: 9; Tamm 2015b); the theory of memory and history in Lotman’s work stems from a more general semiotic theory of culture (Tamm 2019); Halbwachs’s theory of memory is part of a more general theory of social groups. Likewise, Eco’s interpretative model lends itself to the elaboration of a semiotic theory of memory. In particular, by envisioning semiosis and structure together^[4], that is, the analysis of the flow of interpretants and of the local conditions of interpretation, Eco’s semiotics offers us a theoretical perspective and methodological tool for bringing together the double picture of memory-as-containment and memory-as-flow. Since his model has not yet been adequately addressed and integrated – theoretically and methodologically – in memory studies, I will aim to do so in the following three sections.

6 Mnemonic techniques and ars oblivionalis

Eco’s first approach to remembering and forgetting dates back to 1966. At a Symposium on Semiotics and Memory, he presented a paper that would appear in English in 1988, with the title “An ars oblivionalis? Forget it!” (Eco 1988). In this text, Eco deals with memory and oblivion by analysing mnemonic techniques as outlined in Latin memory treatises (the rhetorical art developed for fixing and recalling pieces of information and knowledge). Mnemonic techniques, as semiotics, organise and link two series of elements: the former, acting as a signifier (plane of expression), helps to recall the latter (plane of content). Since “semiotics is by definition a mechanism that presents something to the mind” (Eco 1988: 259), Eco argues that it is impossible to elaborate an ars oblivionalis drawing on the model of a mnemonic technique; if we try to forget something intentionally, the result is always that we make it present and recall it. However, Eco points out that mnemonic techniques do not just tell us what to remember, but also what to forget, because they fix “the distinctive features that have to be recalled in the course of remembering, to the detriment of other features” (Eco 1988: 260). Actually, as Eco points out in a footnote, this process of selection – between what to remember and what to forget – is not specific to mnemonic techniques alone:

interpreting the expression in context means magnifying certain interpretants and narcotizing others, and narcotizing them means removing them provisionally from our competence, at least for the duration of the interpretation taking place . . . I learn something more, basically, by learning something less – that is, by excluding all the other interpretations of the same expression that I could have offered. But such oblivion is transitory; it is a side effect provoked by the interpretive economy. In order to work, a semiotics presupposes the possession of an encyclopedic information. The process of production and of interpretation of texts, by contrast, encourages these passing pseudocancellations. (Eco 1988: 260)

This footnote synthetically introduces – in a text originally conceived in the 1960s – some of the elements of the theory of interpretation that Eco developed in the following decades (Eco 1976 [1975], 1979, 1984). It clarifies that memory- and oblivion-making represent not a semiotic effect limited to a particular discursive technique, but the result of any interpretive process: mnemonic techniques draw on a cultural mechanism that is present in any text and semiotic system.

Finally, in the 2000s, Eco integrated a theory of memory as part of his more general theory of interpretation. Specifically, he did so in the essay “From the Tree to the Labyrinth” (Eco 2014 [2007]: 3–94), which offers the most complete overview of his theoretical work on the encyclopedic model.^[5] Eco articulates the concept of Encyclopedia on two macro-levels (Violi 2015). At a global level, we ideally have all the recorded knowledge that is imagined as acentric and non-hierarchical networks of nodes, as the result of the flow of interpretants: the so-called unlimited semiosis. This is what Eco calls Maximal Encyclopedia. At a local level, texts – and in general, cultural artefacts, social and symbolic practices – activate only some sets of knowledge and information from those virtually available and stored in the Maximal Encyclopedia, leaving the others in abeyance, according to the local conditions of interpretation. The pruning of the Maximal Encyclopedia generates Local Encyclopedias, which are the result of the selection and filtering of knowledge. According to Eco, it is such processes of filtering and selection that produce memory and forgetting, which are – from this perspective – the result of interpretative processes.

It is the combination of these two levels in Eco’s model that allows us to envision remembering and forgetting as the result of the merging of flow and containment. On one hand, the Maximal Encyclopedia focuses on the flow of interpretations that is recorded as networks of cultural units. We could say that the Maximal Encyclopedia records all the versions we produce of the world (and of the worlds), regardless of the whether they are true or false, and of the substances, genres and media we use. On the other hand, the Local Encyclopedias are the result of the filtering of knowledge stored in the Maximal Encyclopedia, within local conditions of use. We could say that they are the local versions we produce, and give shape to, of a limited portion of the world. This capacity of the interpretative model, namely to take into consideration both dynamics, makes it particularly effective in grasping the double and puzzling nature of memory. In the next two sections, I will analyse these two levels – the global and the local – more in depth, and their relevance for the elaboration of an interpretative theory of memory.

7 Maximal Encyclopedia and storage memory

According to Eco, the Maximal Encyclopedia “represents the sum total of everything that was ever thought or said, or at least of everything that could in theory be discovered, to the extent to which it has been expressed through a series of materially identifiable interpretants (graffiti, stelae, monuments, manuscripts, books, electronic recordings) – a sort of World Wide Web far richer than the one to which we have access through the Internet” (Eco 2014 [2007]: 70). If, for Aleida Assmann, storage memory is “the memory of memories” (Assmann 2011 [1999]: 134), Eco considers the Encyclopedia “the library of libraries” (Eco 1984): an archive of all existing information, or in other words, a global multimedia inventory of materially recorded knowledge. As Eco points out (1984: 2), the encyclopedic model brings together two concepts: the Peircean notion of unlimited semiosis, and the Model Q introduced by Ross Quillian, in his research on semantic memory (Quillian 1968).

Drawing on Peirce’s philosophy, Eco argues that a sign, in order to be interpreted, needs to be linked to a new sign or to a chain of signs; the latter are “interpretants,” in relation to the previous signs. Such a production and connection of interpretants is potentially endless: what Peirce calls “unlimited semiosis.” The Maximal Encyclopedia records semiosic activity and – as the production and connection of interpretants – is not just “an inventory of knowledge,” as has often been argued: pieces of knowledge and information – which take a material form – are not just recorded but built by the unlimited semiosis, forming nodes of networks in processes of interpretation. Hence, the Maximal Encyclopedia does not simply store knowledge, but it also creates it through connections, allowing us to learn “something more,” in Peirce’s words: pieces of information are recorded only through linkages in the flow of interpretations. Eco calls the nodes of the network “cultural units.” A cultural unit may interpret another cultural unit, the former potentially becoming the interpretant of the latter:

In fact, we can “touch” interpretants (i.e., we can empirically test a cultural unit), for culture continuously translates signs into other signs, and definitions into other definitions, words into icons, icons into ostensive signs, and ostensive signs into new definition, new definitions into propositional functions, propositional functions into exemplifying sentences and so on; in this way it proposes to its members an uninterrupted chain of cultural units composing other cultural units, and thus translating and explaining them. (1976 [1975]: 71)

The Maximal Encyclopedia “as a global representation is only a semiotic postulate, a regulative idea, and takes the format of a multidimensional network that has been described as the Model Q” (Eco 1984: 68). The Model Q is a model of semantic memory elaborated by Quillian (1968), and to which Eco returns many times (in the elaboration of his interpretative theory, Eco 1976 [1975]: 121–125; Eco 1984: 68–70; Eco [2014] 2007: 57–59). The model Q is a multidimensional network in which

any node can be taken as the point of departure or type of a series of other nodes (tokens) that define it (let’s say the point of departure is dog and that this node is defined by its links with animal, quadruped, able to bark, faithful, etc.). Each of the defining terms may in its turn become the type of another series of tokens. For instance, animal could be exemplified by dog, but also by cat, and would include quadruped but also biped; or, if a node cat were to be identified, it would be defined by a number of nodes it shared with the definition of dog, such as animal and quadruped, but it would also refer to nodes like feline, which it shares with tiger, and so on.

A network model implies the definition of every concept (represented by a term) through its interconnection with the universe of all the concepts that interpret it, each of them ready to become the concept interpreted by all the others.

If we were to expand the network of linked nodes ad infinitum, from a concept assumed as type it would be possible to retrace, from the center to the outermost periphery, the entire universe of the other concepts, each of which may in its turn become the center, thereby generating infinite peripheries. (Eco 2014 [2007]: 57)

The Maximal Encyclopedia is an ensemble of unattainable and unmanageable connections, an acentric and non-hierarchical network of pieces of information that Eco imagines, drawing on a model of semantic memory, as a rhizome. Quillian provides a model for explaining how we share general knowledge of the world, but on a level that is still disembodied and uninhabited, and to be actualized in situated interpretative processes. As noted above, the notion of Maximal Encyclopedia is very close to what Aleida Assmann calls “storage memory” and Lotman “informative memory.” However, differently from these two concepts, the Maximal Encyclopedia is a condition of existence for any memory- and oblivion-making process, its necessary virtual background, on which we draw in order to actualize, or not, the networked pieces of knowledge stored in our culture. It is in this sense that the Maximal Encyclopedia is a postulate and a regulative idea: unattainable and not representable, but also “the only means we have of giving an account, not only of the workings of any semiotic system, but also of the life of a given culture as a system of interlocking semiotic systems” (Eco 2014 [2007]: 51).

Eco’s epistemological imagination becomes methodologically operative when we move from the global to the local level; here we can analyse the processes of actualization of the knowledge stored in the Maximal Encyclopedia, which generates memory and oblivion locally.

8 Local encyclopedias and texts as producers of memory and forgetfulness

Eco concludes his essay “From the Tree to the Labyrinth” with a subsection titled “The Text as Producer of Forgetfulness” (part of Section 1.9., on the Encyclopedia’s formats), where he points out that forgetfulness and remembering are produced “at the level of the textual processes themselves” (2014 [2007]: 90). This opens the possibility to apply the model of textual semiotics and cooperative interpretation (Eco 1979), to the analysis of memory-making in order to explain how texts and, in general, cultural artefacts (e.g., monuments, museums, images) and social and symbolic practices (e.g., informal conversations, commemorations, rites, traditions) generate remembering and forgetting by selecting and actualizing only some very limited sets of knowledge from those virtually available. Here again, there is a very strong resemblance between the Maximal Encyclopedia as storage memory and the Local Encyclopedia as functional memory. In Aleida Assmann’s words, storage memory is “the ‘amorphous mass’ of unused and unincorporated memories that surround the functional memory like a halo” (2011 [1999]: 125); in functional memories, elements stored in material representations are “invested with perspective and relevance,” entering “into connections, configurations, compositions of meaning” (2011 [1999]: 127). Eco’s model of cooperative interpretation helps us to understand how this movement between the Maximal Encyclopedia – as virtual background and storage memory, the halo of any interpretative process – and the Local Encyclopedia – as activated knowledge, invested with relevance – is empirically possible. Compositions of meaning take shape and can be analysed if we identify those structures that act as selectors and organisers of knowledge, generating memory and oblivion at the level of interpretation.

I will not analyse, in detail, the cooperative model as elaborated by Eco (1979: 14) on this occasion, but I will discuss two key concepts, from which many others have been drawn: context and frame.

8.1 Context

As Eco had already pointed out in his analysis of mnemonic technique and ars oblivionalis, the meaning of an expression “is a packet, potentially a quite vast bundle of instructions for interpreting the expression in diverse contexts” (1988: 260; my emphasis). What is relevant and pertinent is recalled and then remembered, what is not relevant within a given context is left in the background, virtually available but removed in the act of interpretation. According to Eco, “context” is actually an umbrella term that can be articulated at three levels: circumstance, context and co-text.

First of all, circumstances are the settings in which a given expression occurs. Circumstantial selections act as such “when the addressee connects the received expression with the act of utterance” and with the social and cultural characteristic of the environment (Eco 1979: 19). Eco (1976 [1975]: 114) offers the example of a red flag (Figure 1).

Figure 1:

Eco’s formalisation of circumstantial selections.

A red flag (as expression) has different potential meanings (“red flag,” in the schema, as a cultural unit condensing in the Encyclopedia all its potential meanings and connections), ranging from political symbolism (i.e., communism) to more practical uses (e.g., to indicate a danger): thus, a red flag used on a motorway calls for “caution”; on a railroad it means “stop”; during a political rally it symbolises “communism.” Hence, circumstances (circ, in the schema) link a semiotic system – the situation – with another semiotic system, in this case, the different uses of a red flag. In doing so, they select some among all the possible meanings, while removing or even overwriting others.

As we will see in the case of Fossoli, certain memorial practices – commemorative days, in particular – work semiotically as circumstances, selecting sets of knowledge and instructing us about what, when and how to remember and forget. Thus, memorial days dedicated to Italy’s Liberation from Nazi-fascism, Holocaust Remembrance Day or National Memorial Day activate different sets of knowledge on the former concentration camp from those virtually available, provisionally removing the others.

Secondly, context – in a stricter sense, for Eco – represents “the coded abstract possibilities of meeting a given term in connection with other terms belonging to the same systems” (Eco 1979: 19). For example, in a text that tells the story of a journey on a motorway, the expression “red flag” will probably appear together with the word “accident,” but not with the word “demonstration”. Methodologically, circumstances are different from contexts because the former connect different semiotic systems with one another (e.g., the verbal or the visual is linked to a social practice), while the latter are about possible connections within the same semiotic systems, substances or media. Circumstances are particularly relevant when we analyse settings, like rituals, commemorations, spaces of memory or conversations; they put an element of a semiotic system in relation to a setting of use. Contexts are, instead, more relevant when we work within the same semiotic system.

However, circumstances can be transformed into contexts; in a text, circumstances “are verbally expressed and even external circumstances are linguistically described” (Eco 1979: 19). Indeed, textual genres and discursive domains are ways to code social circumstances and symbolic practices, transforming them into contexts and allowing the actualization – or the temporary cancelation – of pieces of information stored in the Maximal Encyclopedia. Returning to Eco’s example, a driver’s handbook textually describes the circumstances in which we might see a red flag and what this means. Likewise, a political song, a poem or a pamphlet also transforms a circumstance into verbal terms – e.g., the description of a political rally or uprising – and then into a context, helping us to immediately disambiguate the meaning of the expression “red flag.”

Contexts and, in particular, the broadening or narrowing down – over time and space – of the possible situations in which an element (a picture, a text, a slogan, etc.) circulate and can be used may explain and help analyse the changing meanings in the remediations of the past (Erll and Rigney 2009). A highly significant example of remediation and reconfiguration of meanings that can be analysed in terms of a contextual change is the iconic image of the so-called “Warsaw ghetto boy” (Rousseau 2009), an icon of the Holocaust. The image’s extraction from its original context, namely the Stroop Report (a Nazi collection intended also as a war trophy for Heinrich Himmler), and its adaptation and use across different times, contexts and cultures, and for multiple purposes (from book covers and posters to art installations, commemorations and political demonstrations), allowed for the creation of new configurations of meaning, which were opposed to the original one (to celebrate the genocide’s success). These new meanings are the result of the activation of just some among all the meanings that the picture has generated and accumulated over time, recorded in the Maximal Encyclopedia as networks of potential connections of cultural units, and according to the new contexts of use: from the Nazi’s use of the image as a celebration of genocide to its becoming an icon for remembering the Nazi extermination of the Jewish people, up to its recent use in the context of the Israeli-Palestinian conflict (Rothberg 2011).

The third articulation is co-text. Through this term, Eco describes the discursive surrounding of an element: that which accompanies it. Co-texts activate semantic threads within the text (what Eco calls topics), also allowing for the activation of frames.^[6] As we will see in the case study further ahead, when the name of Fossoli co-occurs with the name “Auschwitz,” we activate the topic “Holocaust” and the ideological frame of the struggle against Nazi-fascism. However, when it co-occurs with “Hiroshima,” “Basovizza” or “Dresda,” (Dresden) the text is talking about “Evil,” often temporarily overwriting the specific historical identity of the perpetrators, and constructing the memory of a supposed universal and endless struggle between Good and Evil.

9 The case of the former concentration camp of Fossoli

In this last section, I will describe a case study that will serve as an empirical application of an interpretative theory of memory. I have chosen the case of the former Fossoli concentration camp in Italy (Herr 2016; Luppi and Tamassia 2017), whose meaning in Italian culture is complex, multilayered and highly significant, because it shows the dynamics of memory and oblivion at play.

Located almost at the centre of the Po Valley in the north of Italy, Fossoli is a hamlet, part of the small town of Carpi, in the province of Modena. Its area was used for different purposes during World War II. First of all, it acted as a prison camp for British (and Commonwealth) soldiers in 1942–1943, managed by the Italian Royal Army; second, it was a concentration camp for Jewish people and political opponents, during the Italian Social Republic led by fascists and Nazis. In particular, from the end of 1943–1944, it served as a transit camp for a third (2840 persons) of the Jews that were deported from Italy to death camps, and for 2700 political opponents to the Nazis and fascists; in 1944, finally, it acted as a camp for civilians (not necessarily to be deported) and for people to be deported as forced labourers in the German military industry. After the war, the camp was used as a prison and as a centre for “undesired people,” by the Allies at first and subsequently by the newly founded Italian Republic, initially for fascists, and then for foreigners and displaced persons (1945–1947); as a Catholic community (called Nomadelfia), for supporting orphans (1947–1952); and as a refugee camp for displaced Italians from Istria and from the territories around the Istrian peninsula (1954–1970), which became part of the former Socialist Federal Republic of Yugoslavia after the war.^[7]

In the second part of the twentieth century, the word “Fossoli” came to designate the concentration camp site located in the hamlet. It was no longer just a toponym, but started to designate specific events and became part of different narratives: the Resistance and anti-fascism (being a prison for many partisans, who were deported and often also died there); the Holocaust and the history of Nazi-fascism (being a transit camp in the logistics of the deportation of Jews); the history of Italian contemporary Catholicism (having hosted Nomadelfia); the history of the Italian Eastern border and communist regimes (recalling the vicissitudes of Istrians and Dalmatians).

To study the memory of Fossoli by using the interpretative model described in this article means to study “Fossoli” as “a memory figure” (Assmann 2011 [1992]: 180): a shorthand that condenses the past into succinct and transportable mnemonic forms. Jan Assmann’s idea of a “memory figure” is very close to what Eco called “cultural unit.” We can thus consider “Fossoli” as a vector that traverses various circumstances, contexts, co-texts and frames, with all its virtual meanings that, however, are locally selected and actualized according to the semiotic conditions of use, determining different configurations of meaning. Hence, to study the memory of Fossoli means analysing the conditions of the local activation and deactivation of the different pieces of knowledge globally recorded in the Encyclopedia. In other words, to study the memory of Fossoli from an interpretative perspective allows us to see how structures shape the flow of semiosis, by selecting and organising knowledge.

In order to track “Fossoli” as a memory figure and cultural unit, I consulted the digital and analog archives of four Italian newspapers, from 1970 (when the camp, as such, was closed) to 2010. I selected four among the most important and relevant – in terms of circulation – Italian newspapers: Corriere della Sera, La Repubblica, La Stampa and L’Unità (the latter is particularly useful for its political connotation, being the newspaper of the former Italian Communist Party). I identified newspaper reports where the word “Fossoli” was present, thus creating a corpus of about 450 newspaper reports dealing with Fossoli at different degrees of centrality: from just mentioning it to in-depth narrations. Newspaper reports allow us to map the multimedia cultural production around Fossoli; book, movie and theater reviews help us to understand where, when and how Fossoli is narrated. Within this cultural production, three elements gain particular prominence: the opening of the Museum and Monument to the Political and Racial Deportee in Carpi, in 1973; the works of the Jewish survivor and writer Primo Levi who, captured as a partisan, passed through Fossoli before being deported to Auschwitz; commemorative dates included in the calendar.

In the remainder of this last section, I will first show how the action and change of circumstances, contexts and frames instruct the reader to activate and deactivate pieces of knowledge, thus generating acts of remembering and forgetting, which are then considered as effects of interpretation. Next, I will show how this material can be organized in order to study both the afterlife of Fossoli and its memoryscape.

9.1 Contextual and circumstantial selections: commemorative dates and ideological frames

As described above, commemorative dates can act as circumstances that instruct us on how, when and what to remember and forget. Since the 2000s, Fossoli has been linked to three main national commemorative dates: 27 January, Holocaust Remembrance Day (established in Italy in 2000); 10 February, National Memorial Day of the Exiles and Foibe (established in 2004); 25 April, Liberation Day or Anniversary of the Resistance (established in 1946). When media talk about “Fossoli” during these three commemorative days, the interpretants that are immediately activated are those linked to three phenomena: the Holocaust, the Resistance and anti-communism.

On Holocaust Memory Day, Fossoli is remembered because of the Jewish victims that transited through the camp (1943–1944). On National Memorial Day, by contrast, Fossoli is remembered as the refugee camp that hosted Italians from Istria and Dalmatia (1954–1970). Finally, on 25 April, Fossoli is remembered as a place linked to the Resistance, where many anti-fascists were imprisoned and some found their death. Commemorative days become contextual selectors when we analyze, as in this case, press coverage: dates as reported in newspapers and co-texts textualize the circumstance, thus instructing the interpretation, that is, indicating which sets of knowledge to activate while putting others in abeyance, that is, removing and temporarily forgetting them. See Figure 2 for Eco’s notation.

Figure 2:

Circumstantial selections in the case of Fossoli.

The institution of the different commemorative days follows the ideological frames that have shaped Italian public memory over time: the Resistance from the 1940s to the 1980s, the Holocaust since the 1980s, and the change in politics of memory after the end of the Cold War (with periods of transitions, overlapping and competition between the different ideological frames). This timeline is reflected in co-texts that accompany “Fossoli” in newspapers.

In the 1970s, “Fossoli” co-occurs with names of other places like “Piazzale Loreto,” a square in Milan where partisans were executed during World War II, or with the names of other Italian prisons from where political opponents were deported, thus activating knowledge related to the Liberation struggle.^[8] At the end of the 1970s, Fossoli was linked to a media event: the American TV series Holocaust.^[9] In this case, “Fossoli” co-occurred with the word “Holocaust” but also with “Auschwitz,” clearly activating the layer of meaning specifically linked to the Jews’ persecution.^[10] From the 2000s onwards, “Fossoli” occurred also with words like “Basovizza” or “foibe,” activating a different set of historical knowledge – the history of communist regimes in East Europe and human rights violations – and connecting it with Fossoli. This new co-text, linked to the new commemorative circumstance, is explained by a change in politics of memory and also by a conflict over memory (Foot 2009). In fact, a right-wing government instituted the Memorial Day for the foibe in 2004, with an anti-communist aim, namely to contrast the memory of the Resistance, in which the symbolic legacy of communist militants is particularly important. This ideological frame generated a clash of memory narratives, making the two sets of knowledge – the Resistance and the history of refugees and Foibe at the eastern boarder – conflicting and mutually exclusive: on the one hand, the representation of communist partisans, imprisoned and even killed in Fossoli, therefore acting as heroes and victims within the narrative of the Resistance (Subject or Object of value in the actantial model), clearly clashes with the representation of communists as perpetrators (anti-Subject), when we remember refugees from the eastern border, on the other hand.

A further shift occurred at the end of the 2000s. In 2009, the Mayor of Rome – a militant of a post-fascist party – announced his intention to visit “Fossoli” as well as the “Foibe” and “Hiroshima.”^[11] He thus created an unprecedented link that tried to gloss over the difference between the Holocaust, communist regimes and the military struggle against Nazi-fascism and its allies, combining everything in a general and ahistorical category of absolute Evil. The specific, historical identities of perpetrators were thus removed from the process of interpretation, by way of excluding the recognition of fascist historical responsibility. Hence, linking together different words and interpretants, co-texts activate ideological frames and trigger the removal of knowledge and information about the past on ideological bases.

The shift in ideological frameworks emerges particularly well if we look at two cultural elements strictly linked to the Fossoli camp: the Museum and Monument to the Political and Racial Deportee, and the figure of Primo Levi. Newspaper reports in the 1970s represent Fossoli as a place connected to the narrative of the Resistance. The project for the Museum’s construction is described as an act of reparation, for a region that perceives the presence of the camp as an “insult” to the memory of the local Resistance heroes.^[12] In 1973, the inauguration of the Museum, by the President of the Italian Republic, was the apex of a presidential pilgrimage to the locations of the Resistance in the north of Italy. On the walls of the museum, visitors can read extracts from the letters written by partisans of the European Resistance, who had been sentenced to death by Nazis and fascists. These letters still represent the Museum’s narrative backbone. Although newspapers describe the presence of the Italian and local Jewish community with different degrees of relevance, the Museum and its inauguration are narrated and understood within the context of the memory of the partisan struggle; the Jewish peculiarity of the deportation is not fully acknowledged. Although the Museum-Monument has not undergone any significant changes since its inauguration, in the 1990s the museum was defined “the first Museum of the Holocaust” in Europe,^[13] an utterly anachronistic definition if we look at the project and its inauguration. In the two different descriptions of the Museum-Monument, in the 1970s and 1990s, we clearly see the shift in grand narrative that is used to make sense of the place: from the struggle between Nazi-fascist dictatorship and the Resistance, centred on the heroic figure of the partisan, to the narrative of historical injustice and victimhood during the Holocaust.

This process can also be seen in action in the representation of the figure of Primo Levi. In 1974, Corriere della Sera published an article on the occasion of the 29th anniversary of the liberation of Mauthausen, also remembering Nazi camps in general. In the article, Levi is introduced as a partisan and an anti-fascist; his Jewish origin is barely mentioned in the biographical note.^[14] His deportation to Auschwitz via Fossoli is understood within the frame of the partisan struggle, rather than being linked to his being a Jewish survivor of the persecution. A decade later, by contrast, Levi’s identity is centred on his Jewishness, while his partisan identity is downplayed and put in a very marginal narrative position, if not ignored.^[15] The radical change in the identity of the Museum-Monument, and of key figures such as Primo Levi, can be read as an effect of the ideological frame that also guides the construction of the memory of Fossoli: from the Resistance as a national founding narrative of the Italian Republic to the transnational framework of the Holocaust. As we will see next, such changes affect narrative structures and roles in narrations (e.g., those of heroes and victims) as well.

10 Conclusion

In order to link the past with the present and connect different spaces and generations, cultural memory installs itself locally yet continuously breaks away from and out of its provisional frameworks in order to emerge somewhere else. When analyzing cultural memory, we are asked to acknowledge this double nature of memory.

In this article, I have therefore argued that we must overcome the counterproductive, dichotomic vision of memory-making as either containment or flow, by drawing on cultural and interpretative semiotics. Eco and Lotman’s understandings of culture – as an assemblage of correlated systems that record, organize, activate and deactivate knowledge and information – offer memory studies scholars a perspective through which to envision memory-making both as movement and as a local mise en forme.

Focusing in particular on Eco’s model and on an interpretative theory of memory, I have argued for a conceptualization of remembering and forgetting as a result of the local selection and organization of knowledge. This implies that we should: 1) reconstruct the production of the flow of interpretants (e.g., texts, images, practices, narratives) that are produced around a given cultural unit – in other words, we must follow semiosis; 2) analyze the structures that locally drive semiosis, within genres, substances and media, by instructing the selection of knowledge and generating remembering and forgetting that shape our knowledge of the past; 3) study the correlations between the n versions of the past we thus produce, coherently or contradictorily, synchronically (memoryscape) or over time (afterlife). We should do so by drawing on and connecting different semiotic substances, media, genres and domains – and their mutual, conflicting or consistent, interlocking – each with its own way of forming and shaping the world.